Hydrometry: Design Manual Volume-4 2003

Government of India & Government of The Netherlands
DHV CONSULTANTS &
DELFT HYDRAULICS with
HALCROW, TAHAL, CES,
ORG & JPS
VOLUME 4
HYDROMETRY
DESIGN MANUAL

Design Manual – Hydrometry (SW) Volume 4
Table of Contents
1 INTRODUCTION 1
1.1 GENERAL 1
1.2 DEFINITION OF VARIABLES AND UNITS 2
2 PHYSICS OF RIVER FLOW 5
2.1 GENERAL 5
2.2 CLASSIFICATION OF FLOWS 5
2.3 VELOCITY PROFILES 8
2.4 HYDRAULIC RESISTANCE 10
2.5 UNSTEADY FLOW 14
2.6 BACKWATER CURVES 17
3 HYDROMETRIC NETWORK DESIGN 21
3.1 INTRODUCTION 21
3.2 NETWORK DESIGN CONSIDERATIONS 22
3.2.1 CLASSIFICATION 22
3.2.2 MINIMUM NETWORKS 22
3.2.3 NETWORKS FOR LARGE RIVER BASINS 23
3.2.4 NETWORKS FOR SMALL RIVER BASINS 23
3.2.5 NETWORKS FOR DELTAS AND COASTAL FLOODPLAINS 23
3.2.6 REPRESENTATIVE BASINS 24
3.2.7 SUSTAINABILITY 24
3.2.8 DUPLICATION AVOIDANCE 24
3.2.9 PERIODIC RE-EVALUATION 24
3.3 NETWORK DENSITY 24
3.3.1 WMO RECOMMENDATIONS 24
3.3.2 PRIORITISATION SYSTEM 25
3.3.3 STATISTICAL AND MATHEMATICAL OPTIMISATION 26
3.4 THE NETWORK DESIGN PROCESS 26
4 SITE SELECTION OF WATER LEVEL AND STREAMFLOW STATIONS 28
4.1 DEFINITION OF OBJECTIVES 28
4.2 DEFINITION OF CONTROLS 28
4.3 SITE SURVEYS 29
4.4 SELECTION OF WATER LEVEL GAUGING SITES 31
4.5 SELECTION OF STREAMFLOW MEASUREMENT SITE 31
5 MEASURING FREQUENCY 34
5.1 GENERAL 34
5.2 STAGE MEASUREMENT FREQUENCY 35
5.3 CURRENT METER MEASUREMENT FREQUENCY 36
6 MEASURING TECHNIQUES 38
6.1 STAGE MEASUREMENT 38
6.1.1 GENERAL 38
6.1.2 VERTICAL STAFF GAUGES 39
6.1.3 INCLINED STAFF OR RAMP GAUGES 42
6.1.4 CREST STAGE GAUGES 43
6.1.5 ELECTRIC TAPE GAUGES 44
6.1.6 FLOAT SYSTEM WITH AUTOGRAPHIC RECORDING 45
6.1.7 FLOAT SYSTEM WITH DIGITAL RECORDING 50
6.1.8 PRESSURE SENSORS (TRANSDUCERS) 52
6.1.9 DATA LOGGERS 59
Hydrometry January 2003 Page i

6.1.10 SELECTION OF STAGE AND WATER LEVEL SENSORS AND
RECORDING EQUIPMENT 63
6.2 INTRODUCTION TO VELOCITY AREA METHOD OF STREAMFLOW
MEASUREMENT 64
6.2.1 BASIC PRINCIPLES 64
6.2.2 FACTORS AFFECTING VELOCITY DISTRIBUTION IN OPEN
CHANNELS 66
6.2.3 DISTRIBUTION OF VELOCITY 69
6.3 FLOAT MEASUREMENT 70
6.3.1 BACKGROUND 70
6.3.2 TYPES OF FLOAT 71
6.3.3 FLOAT MEASUREMENT TECHNIQUE 72
6.3.4 DETERMINATION OF DISCHARGE FROM SURFACE FLOAT
VELOCITY MEASUREMENTS 74
6.4 CURRENT METER STREAMFLOW MEASUREMENT 76
6.4.1 INTRODUCTION 76
6.4.2 ROTATING ELEMENT CURRENT METERS 76
6.4.3 TYPES OF CURRENT METER 78
6.4.4 METHODS OF SUSPENSION AND DEPLOYMENT OF CURRENT
METERS 81
6.4.5 SPACING OF VERTICALS 86
6.4.6 MEASUREMENT OF WIDTH, HORIZONTAL DISTANCE OR POSITION
IN THE HORIZONTAL 88
6.4.7 MEASUREMENT OF DEPTH 91
6.4.8 SKEW EFFECTS 97
6.4.9 METHODS OF ESTIMATING MEAN VELOCITY IN THE VERTICAL 98
6.4.10 LEGITIMATE SHORT CUTS 100
6.4.11 COMPUTATION OF DISCHARGE 100
6.4.12 ERROR ANALYSIS VELOCITY AREA METHOD 101
6.4.13 MINIMISING ERRORS IN CURRENT METER DISCHARGE
MEASUREMENTS 106
6.4.14 ELECTROMAGNETIC CURRENT METER 106
6.5 ACOUSTIC DOPPLER CURRENT PROFILER (ADCP) 107
6.5.2 BASIC PRINCIPLES OF DOPPLER VELOCITY METERS 108
6.5.3 ADCP FUNCTIONING AND COMPONENTS 110
6.5.4 PRACTICAL USE OF THE ADCP 113
6.5.5 COLLECTING DISCHARGE DATA WITH A SONTEK ADP 118
6.5.6 PROCESSING OF SONTEK ADP DATA 120
6.5.7 ADCP - GLOSSARY OF TERMS 123
6.6 SLOPE - AREA METHOD 125
6.6.2 PRINCIPLES OF THE METHOD OF MEASUREMENT 125
6.6.3 ESTIMATION OF VELOCITY HEAD AND DISCHARGE 128
6.6.4 ESTIMATION OF MANNING’S COEFFICIENT 128
6.6.5 COMPOSITE SECTIONS 129
6.6.6 STATE OF FLOW 130
6.7 COMPARISON AND APPLICATION OF DIFFERENT FLOW MEASUREMENT
TECHNIQUES 130
7 EQUIPMENT SPECIFICATIONS 134
7.1 INTRODUCTION 134
7.2 PREPARATION OF STAGE AND STREAM FLOW MEASUREMENT
INSTRUMENTATION AND EQUIPMENT TENDER SPECIFICATIONS 135
7.2.2 SITE SURVEYS AND INSTALLATION PLANNING 135
7.2.3 TENDER PACKAGES 136
Hydrometry January 2003 Page ii

7.2.4 RESOLUTION AND ACCURACY 137
7.2.5 HYDROMETRIC SENSOR AND EQUIPMENT SELECTION CRITERIA 137
7.2.6 INSTRUMENT AND EQUIPMENT SPECIFICATIONS 138
7.2.7 OTHER ITEMS TO BE INCLUDED IN THE SPECIFICATIONS 141
7.3 INTERPRETATION OF EQUIPMENT SPECIFICATION AND ASSESSMENT OF
PRODUCTS AND SUPPLIERS 143
7.3.1 INTERPRETING EQUIPMENT SPECIFICATIONS 143
7.3.2 EVALUATION OF SPECIFICATIONS 143
7.4 EQUIPMENT SPECIFICATIONS 145
8 STATION DESIGN, CONSTRUCTION AND INSTALLATION 147
8.1 DESIGN AND INSTALLATION CRITERIA - WATER LEVEL MONITORING 147
8.1.1 STAFF GAUGES 147
8.1.2 BENCH MARKS 149
8.1.3 STILLING WELLS 149
8.1.4 DWLR - PRESSURE SENSOR TYPE 154
8.1.5 INSTRUMENT PROTECTION AND HOUSINGS 157
8.2 DESIGN AND INSTALLATION CRITERIA - CURRENT METER GAUGING
INSTALLATIONS 158
8.2.1 GENERAL 158
8.2.2 WADING GAUGING SITES 159
8.2.3 BRIDGE GAUGING 159
8.2.4 CABLEWAY GAUGING 160
8.2.5 BOAT GAUGING 165
8.3 SITE OFFICES AND STORES 166
8.3.1 SITE TYPES 166
8.3.2 ASSUMPTIONS 167
8.3.3 GUIDELINES 167
8.3.4 POSSIBLE BUILDING REQUIREMENTS 168
9 REFERENCES 169
Hydrometry January 2003 Page iii

1 INTRODUCTION
1.1 GENERAL
The branch of Geophysics, which deals with the occurrence and movement of water in terms of
quantities and quality on and below the surface of the earth except the oceans, in vapour, liquid or
solid state, is termed Hydrology. For hydrological design and water resources assessment purposes
proper estimates of river flow and river stages are required. Their measurement is the domain of
hydrometry.
Figure 1.1:
Hydrometric station
The measurement of river stages and discharges at the observation stations is dealt with in this
Volume 4 “Hydrometry” of the “Manual on Hydrological Field Measurements and Data Processing”.
This volume on hydrometry includes how measurements are made, with what equipment, where and
when. Volume 4 consists of three parts:
1. Design Manual, in which the basic principles and procedures are put in context
2. Reference Manual, for details on specific topics, and
3. Field Manual, dealing with operational procedures at the observation station.
This part of Volume 4 covers the Design Manual: ‘Hydrometry’. It is set up as follows:
• Chapter 1 deals with definition of quantities and units and unit conversions.
• Some basic hydraulic principles as far as relevant for hydrometry are dealt with in Chapter 2.
• In Chapter 3 the design and optimisation of hydrometric networks are discussed. Network
densities are related to measurement objectives, spatial variation of the phenomena and cost of
installation and operation.
Hydrometry January 2003 Page 1

• Once the network density has been specified the sites for the water levels and discharges have to
be selected. Criteria for site selection are discussed in Chapter 4.
• Next, in Chapter 5 the observation frequency to be applied for the various hydrological quantities
in view of the measurement objectives and temporal variation of the observed processes are
treated.
• The measurement techniques for observation of hydrometric variables and related equipment are
dealt with in Chapter 6.
• Since the buyers of the hydrometric equipment are often neither sufficiently familiar with the exact
functioning of (parts of) the equipment nor with the background of the specifications, remarks on
the equipment specifications have been added in Chapter 7. The equipment specifications proper
are covered in a separate and regularly updated volume: “Equipment Specification Surface
Water”.
• Guidelines on station design and equipment installation are dealt with in Chapter 8.
In the Field Manual operational practices in running the network stations are given. It also includes
field inspections, audits and last but not least, the topic of equipment maintenance and calibration.
Notes
• The content of this part of the manual deals only with hydrometric measurements in the States of
Peninsular India. The equipment discussed is used or appropriate for use in the Hydrological
Information System. Hence, the manual does not provide a complete review of all techniques and
equipment applied elsewhere.
• The procedures dealt with in this manual are conformably to BIS and ISO standards. It is
essential that the procedures described in this manual are closely followed to guarantee a
standardised approach in the entire operation of the Hydrological Information System.
1.2 DEFINITION OF VARIABLES AND UNITS
In this section definitions, symbols and units of relevant quantities and parameters when dealing with
hydrometry are given. The use of standard methods is an important objective in the operation of the
Hydrological Information System (HIS). Standard methods require the use of a coherent system of
units with which variables and parameters are quantified. This section deals with the system of units
used for the measurement of hydrological quantities.
Quantity Symbol Unit Quantity Symbol Unit
Density
Density of water
Density of sediment,
Relative density under water
Pressure
Air pressure
Water pressure
Temperature
Water temperature
Air temperature
Level, depth, area
Water depth
Wetted perimeter
Wetted area
Hydraulic radius
Equilibrium or normal depth
Critical depth
ρ
ρs
Δ=(ρs-ρ)/ρ
pa
p
Tw ,tw
Ta ,ta
y, h
P
A
R
yn, hn
yc, hc
kg.m-3
kg.m-3
[-]
kPa
kPa
oC or K
oC or K
m
m
m2
m
m
m
Head
Velocity head
Pressure head
Energy head
Slope
Slope/gradient (general)
Bottom/bed slope/gradient
Water surface slope/gradient
Energy slope/gradient
Discharge
Flow velocity
Discharge
Discharge per unit width
Characteristic numbers
Reynolds number
Froude number
hv
hp
He
S
S0
Sw
Se
u, v, w
Q
q
Re
Fr
m
m
m
[-]
[-]
[-]
[-]
m.s-1
m3.s-1
m2.s-1
[-]
[-]
Table 1.1: Overview of relevant quantities, symbols and units used in hydrometry

General terms
Control: The physical properties of a channel, natural or artificial, which determine the relationship
between stage and discharge at a location in the channel.
Stable channel: Channel in which the bed and the sides remain sensibly stable over a substantial
period of time in the control reach and in which scour and deposition during the rising and
falling floods is inappreciable.
Unstable channel: Channel in which there is frequently and significantly changing control.
Reach: A length of open channel between two defined cross-sections
Invert: The lowest part of the cross-section of a natural or artificial channel.
Wetted perimeter, P [m]: The wetted boundary of an open channel at a specified section.
Cross-section of stream, A [m2]: A specified section of the stream normal to the direction of flow
bounded by the wetted perimeter and the free water surface.
Hydraulic radius, R [m]: The quotient of the wetted cross-sectional area and the wetted perimeter.
Level, depth and gradient
Stage, y, h [m]: Height of water surface of a stream, river, lake or reservoir at the measuring point
above an established datum plane.
Gauge height, h [m]: Water surface elevation relative to the gauge datum.
Water depth D, h [m]: Vertical distance between water surface and river bottom.
Normal/equilibrium depth, [m]: Flow depth under steady, uniform flow conditions.
Critical depth, [m]: The depth of flow when the flow is critical (Fr = 1), see Chapter 2.
Gauge: The device installed at the gauging station for measuring the level of the water surface
relative to datum. If the gauge is linked to a standard system of levels then the gauge is a
reference gauge.
Water level recorder: A device which records automatically, either continuously or at frequent time
intervals, the water level as sensed by a float, a pressure transducer, a gas bubbler, acoustic
device, etc.
Stilling well: A well connected to the main stream in such a way as to permit the measurement of
the stage in relatively still water.
Surface slope: The difference in elevation of the surface of the stream per unit horizontal distance
measured in the direction of flow.
Bed/bottom slope: The difference in elevation of the bed per unit horizontal distance measured in
the direction of flow.
Backwater curve: The profile of the water surface upstream when its surface slope is generally less
than the bed slope. The backwater curve occurs upstream of an obstruction or confluence.
Draw-down curve: The profile of the water surface when its surface slope exceeds the bed slope.
Afflux: The rise in water level immediately upstream of, and due to, an obstruction.
Elevation/potential head, [m]: The height of any particle of water above a specified datum (potential
energy per unit of weight relative to a horizontal datum).

Pressure head, [m]: Height of liquid in a column corresponding to the weight of the liquid per unit
area.
Piezometric head, [m]: Sum of elevation head and pressure head, or above a datum, the total head
at any cross-section minus the velocity head at that cross-section.
Velocity head, [m]: The head obtained by dividing the square of the velocity by twice the acceleration
due to gravity. In applying the mean velocity in the cross-section, a correction factor is to be
applied for non-uniformity of the velocity profile in the cross-section.
Total energy head, [m]: The sum of the elevation of the free water surface above a horizontal datum
of a section, and the velocity head.
Specific energy, [m]: The sum of the elevation of the free water surface above the bed, and the
velocity head.
Energy gradient, [-]: The difference in total energy head per unit horizontal distance in the direction
of flow.
Stage-discharge relation: A curve, table or function, which expresses the relation between the stage
and the discharge in an open channel at a given cross-section for a given condition of flow
(rising, steady or falling)
Flow and flow types
Discharge, [m3/s]: Volume of liquid/water flowing through a cross-section per unit of time.
Velocity, [m/s]: Rate of movement past a point in a specified direction.
Laminar flow: Type of flow mainly determined by viscosity Re < 500, see Chapter 2
Turbulent flow: Type of flow which is hardly determined by viscosity: Re > 2000, see Chapter 2.
Sub-critical flow: The flow in which the Froude number is less than unity and surface disturbances
can travel upstream, see Chapter 2.
Super-critical flow: The flow in which the Froude number is greater than unity and surface
disturbances will not travel in upstream direction, see Chapter 2.
Critical flow: The flow at which the total energy head is at minimum for a given discharge; under this
condition the Froude number will be equal to unity and surface disturbances will not travel in
upstream direction, see Chapter 2.
Steady flow: Flow in which the depth and velocity remain constant with respect to time, see
Chapter 2.
Uniform flow: Flow in which the depth and velocity remain constant with respect to distance, see
Chapter 2.
Friction, drag: Boundary shear resistance, which opposes the flow of water.
Friction coefficient: A coefficient used to calculate the energy gradient caused by friction.
Rugosity coefficient: A coefficient linked with the boundary roughness and the geometric
characteristics of the channel used in the open channel flow formulae, like Chezy coefficent,
Manning’s coefficient, etc.
Hydraulic jump: Sudden change of flow from super-critical flow to sub-critical flow.

2 PHYSICS OF RIVER FLOW
2.1 GENERAL
In this chapter an overview is given of some relevant subjects of hydraulics, including:
• classification of flows: laminar versus turbulent flow, sub-critical, critical and supercritical flow and
steady and unsteady flow, varying gradually or rapidly,
• flow velocity profiles for laminar and turbulent flow conditions,
• hydraulic roughness,
• unsteady flow features, and
• backwater computations.
Detailed derivations of the flow equations are provided in Volume 4, Reference Manual, Hydrometry.
2.2 CLASSIFICATION OF FLOWS
Flows in rivers are classified according to the forces acting on a mass of fluid. These are:
gravity Fg = M.g = ρL3g (M = mass; g = gravitational acceleration; ρ = density; L = length)
pressure Fp = p.A = p L2 (p = pressure; A = area)
viscosity Fν = τ.A = ρν v L (τ = shear stress; ν = kinematic viscosity; v = velocity)
surface tension Fσ = σ.L =σL (σ = surface tension)
elasticity Fe = K.A =KL2 (K = bulk modulus of elasticity)
inertia Fi =M.a = ρv2L2 (a = acceleration)
Generally, one of these forces predominates. The inertial force is always present. To characterise the
physical phenomena, the forces are compared with the inertial force leading to characteristic
numbers. For river flow or open channel flow the Reynold(s) number and Froude number are of
importance.
Reynolds number
The Reynolds number Re compares the viscous force with the inertial force:
vL
ν
=
ρ
2 2
v L
ρν
F
i
≡ =
F
ν
vL
Re
For river flow the flow depth h is taken as the characteristic length L: so L→h. Hence, it follow from
(2.1):
≡ vh
ν
Re
(2.1)
(2.2)

The Reynolds number distinguishes between laminar and turbulent flow:
• laminar flow: Re < 600
• transitional flow: 600 ≤ Re < 2000
• turbulent flow: Re > 2000
Laminar flow is best described as thin sheets of water (laminae) moving in straight lines parallel to
each other, although the velocities of one sheet may not be the same as the one beside it. In this
situation the viscosity is very strong relative to the inertia forces. Viscosity is the resistance of
movement of one layer of fluid to another. Very simply it is a measure of a liquid’s “stickiness”. In a
turbulent flow situation, the path of the fluid particles is no longer straight as the viscous forces are
weak relative to the inertial forces. Therefore flows are sinuous and intertwining with each other so
that thorough mixing takes place. In turbulent flow there are continuous variations in velocity (and
pressure) at every point, so only laminar flow can be considered steady. Turbulent flow is only steady
if the average velocity and pressure remain constant over a reasonable time period.
Since the viscosity of water is about 10-6 m2/s at a temperature of 20 oC it is observed from the
Reynolds number that in nearly all cases river flow is turbulent; only sheet flow with very low velocities
will behave as laminar flow. The fact that river flow is turbulent has consequences for measurement of
stage and of flow velocities. A real instantaneous value will give insufficient information about the
state of flow; time averaged values over periods of 0.5 to several minutes have to be considered
instead.
The Froude number Fr, which compares the gravity force with the inertial force:
F
Froude number
ρ
2 = i =
=
F
g
The Froude number reads with L replaced by the flow depth h (or for a channel with non-uniform
cross-section: cross-sectional area/stream width at the surface):
v
gh
Fr
Fr ≡
v
2
gL
2 2
v L
3
ρ
L g
(2.3)
(2.4)
The Froude number compares the celerity of dynamic waves √(gh) with the flow velocity v:
• sub-critical flow: Fr < 1 flow is slow
• critical flow: Fr = 1 flow has unique depth hc = critical flow depth
• supercritical flow: Fr > 1 flow is fast
The specific energy of the flow in a particular cross-section (h + v2/2g) is at a minimum for one
particular depth, called the critical depth hc. For a particular discharge there can only one depth be
critical. Hence, when the flow is critical, there is a unique relation between stage at discharge. Of this
feature use is made of in flow measuring structures. Critical flow is obtained in the transition from a
mildly sloped channel where the flow is sub-critical to a steep channel with very high flow velocities,
where the flow is super-critical. As is observed from the definition of the Froude number in natural
rivers where gauging takes place often the condition Fr << 1 applies, so one is generally dealing with
sub-critical flow.

Flow classification on temporal and spatial variation of flow velocity and depth.
Classification of open channel flow can also be based on the temporal and spatial variation of the
mean flow velocity v and mean flow depth h: v = v(x,t) and h = h(x,t) as shown in Table 2.1 (see also
Figure 2.1):
Steady Flow: Depth of flow does not change with respect to the time period under consideration.
Unsteady Flow: Depth of flow is constantly changing within the time period.
Uniform Flow: Depth of flow does not change with distance under consideration along the
channel.
Non-Uniform Flow: Depth varies with distance.
Gradually Varied Flow: Depth of flow varies very little over a large distance of channel.
Rapidly Varied Flow: Depth of flow changes rapidly over a comparatively short distance, e.g. in a
hydraulic jump
Gradually varied flow occurs in most gradually sloping river systems in India. Rapidly varied flow
occurs at such features as weirs and waterfalls.
Figure 2.1:
Flow classification based on
temporal and spatial variation of
flow velocity and flow depth
OPEN CHANNEL FLOW
STEADY FLOW UNSTEADY FLOW
TYPES
OF
FLOW
UNIFORM FLOW NON-UNIFORM FLOW
GRADUALLY
VARIED FLOW
RAPIDLY
VARIED FLOW
Flow condition |∂v/∂x| |∂v/∂t| |∂h/∂x| |∂h/∂t|
Steady flow 0 0
Uniform flow 0 0 0 0
Non-uniform or varied flow > 0 0 > 0 0
Gradually varied flow small 0 small 0
Rapidly varied flow large 0 large 0
Unsteady flow > 0 > 0
Table 2.1: Classification of flows based on temporal and spatial variation of flow depth

2.3 VELOCITY PROFILES
Consider steady uniform flow. Then, the streamlines are parallel to the riverbed, so bed slope S0 =
water surface slope Sw = energy slope Se, pressure distribution is hydrostatic and accelerations are
zero. The velocity profile is then obtained from a balance of forces in flow direction and a relation
between shear stress and velocity (see Volume 4, Reference Manual).
Laminar flow
In case of laminar flow the velocity profile is parabolic and reads:
1
v(y) 0 − 2
y )
2
(hy
gS
ν
=
where: v(y) = flow velocity at distance y from river bed
By integration over the depth of flow for the average flow velocityv it follows:
gS
h = flow depth
S0 = river bed slope
ν = kinematic viscosity
g = gravitational acceleration
0 h2
3
v
ν
=
Note thatv :: S0, which is characteristic for laminar flow. By comparison of (2.5) with (2.6) it is
observed that the average flow velocity is equal to the velocity at a depth y = (1-1/3√3)h ≈ 0.42h.
Turbulent flow
In case of turbulent flow close to the bottom a very thin laminar sub-layer of depth ‘δ’ exists where the
velocity profile varies linearly with depth. Above the sub-layer the velocity profile is logarithmic, which
is characteristic for fully developed turbulent flow (see Figure 2.2). It is customary to use the shear
velocity u∗ in the expressions for the velocity profiles, which is defined by:
0
τ
u 0 = ghS
ρ
∗ =
where τ0 = bottom shear stress; τ0 = ρghS0
The velocity profiles read:
• In the laminar sub-layer: 0 ≤ y ≤ δ:
y
2
ν
u
v(y)
= ∗
• Above the laminar sub-layer: y > δ:

 
  
u
= ∗
κ
y
y0
ln
v(y)
(2.5)
(2.6)
(2.7)
(2.8)

In equation (2.8) κ = the Von Karman constant, with κ ≈ 0.4, and y0 is the value of y for which the
velocity becomes zero according to the logarithmic profile: v(y0) = 0. The linear and the logarithmic
profile intersect at y = δ. The thickness of the laminar sub-layer is given by:
ν
u
∗
δ =
11.6
(2.9)
In stead of the abrupt change from a linear to a logarithmic velocity profile there is a transition zone
extending from 0.5δ < y < 3δ (i.e. 5ν/u∗ < y < 30ν/u∗).
Figure 2.2:
Velocity profile near
bottom
Transition
zone
Fully
developed
turbulent flow
Laminar
sub-layer
y
δ
v(y)
y0
30ν/u∗
Logarithmic profile
Transitional profile
Linear profile
11.6ν/u∗
5ν/u∗
For common values of h and S0 the thickness of the laminar sub-layer δ << 1 mm. Hence, the average
velocity can safely be derived from equation (2.8) and reads:
(2.10)
h
= ∗ 0
∗

<<  

 
≈

 

 

+ −  

 
v 0
e.y
In
u
y
h
1
h
y
In
u
o
0
κ
κ
Following observations can be made:
because y h
• v :: u∗, so v :: (S0)1/2 and not proportional with S0 like for laminar flow
• v(y) = v for y = h/e = 0.368 h
In equation (2.10) still y0 has to be determined. Its value depends on the roughness of the bottom,
which is characterised by the equivalent sand roughness ks. According to Nikuradse, a bed with
roughness ks produces the same resistance as a flat bed covered with fixed, uniform, closely packed
sand grains with diameter ks. Now the following bed/wall conditions apply:
if ks < 0.3δ, then the bed is hydraulically smooth, and: y0 ≈ δ/117 (2.11)
if ks > 6δ, then the bed is hydraulically rough, and: y0 ≈ ks/32 (2.12)

Combining (2.11) and (2.12) with (2.10) the velocity profiles and average velocities become:
• for a smooth boundary (ks < 0.3δ):
117y


u
∗
ln
κ δ


=
v(y)
u
∗
12h
ln
κ δ
=




/ 3.5
v
• for a rough boundary (ks > 6δ):
32y
s
k

 
u
∗

 
=
v(y)
u
∗
κ
=

 

 
κ
ln
12h
s
k
ln
v
• for the transition between smooth and rough 0.3δ < ks < 6δ the average velocity follows from:
12h
κ + δ
k / 3.5

 
ln
g
12h
= ∗
κ + δ
k / 3.5

=  
s s
0
12h
+ δ
k / 3.5
s
hS or :
0
hS

 
ln
u
v

 
=
v 18 log

 

 
• The above formulae are valid for wide channels. For other cross-sections h has to be replaced by
• In view of the small value of δ in fairly all natural conditions the bed can be considered as
2.4 HYDRAULIC RESISTANCE
Generally two flow equations are in use:
Chezy: v = C(RS)1/2 (2.18)
Manning: v = 1/n R2/3S1/2 (2.19)
where: C = Chezy coefficient [m1/2.s-1]
Using equation (2.17) and replacing flow depth h by the hydraulic radius R and combining the
expression with (2.18) White-Colebrook’s formula for hydraulic resistance is obtained:
C
the hydraulic radius R.
hydraulically rough. Hence, the equations (2.15) and (2.16) generally apply in practice.
n = Manning’s n-value for hydraulic roughness [m-1/3.s]
R
Note:
ks
=
12
+

 

 
18
δ / .
3 5
log
(2.13)
(2.14)
(2.15)
(2.16)
(2.17)
(2.20)

where the denominator in (2.20) takes on the following values:
• For hydraulically smooth bed ks << δ, hence ks+ δ/3.5 ≈ δ/3.5
• For hydraulically rough bed ks >> δ, hence ks+ δ/3.5 ≈ ks.
Strickler proposed the following expression for C:
C
R
k s
=

 

 
25
1/6
Equations (2.20) and (2.21) are almost identical in the range 40 < C < 70. Williamson (1951) found for
concrete tubes the coefficient to be 26.4 instead of 25 for 7.5 < R/ks < 1500.
Combining (2.21) with (2.18) and comparing the result with (2.19) one obtains:
C
1 / 6 1 /
6
R
n
R
k s
= =

 

 
25
Hence the following approximate relation between Manning’s n and Nikuradse’s ks-value exists:
n
1 /
6
k
s k
. /
1 6
s = =
25
0 04
The advantage of the use of ks over n is its dimension [m]. The size of bed unevenness can be
translated into a value for ks (see below). This is at least true for the riverbed. For floodplain
roughness with bushes etc. the relation between unevenness and ks is less apparent.
Some practical relations for ks
According to van Rijn (1984) for an alluvial bed the following values apply for the equivalent sand
roughness ks:
• For a flat sandbed and gravelbed it follows respectively:
k 3D (sandbed) k D (gravelbed) s 90 s 90 ≈ ≈
• For a dune/ripple covered bed (see Figure 2.3)
H
))
L
k 1.1H(1 exp( 25 s ≈ − −
(2.21)
(2.22)
(2.23)
(2.24)
where: D90 = characteristic grain size diameter (90% is finer)
H = dune/ripple height
L = dune/ripple length
H/L = dune/ripple steepness

Figure 2.3:
Dune/ripple dimension
parameters
Note:
Definition of dune/ripple dimension parameters
Water table
Dune covered river bed
L
h
v
H
• For a flat sandbed values for ks in the range of 1 to 10 D90 were found with a median value 3D90
• For steep dunes/ripples H/L is typically 0.1, then ks = H, i.e. sand roughness about equal to the
dune height (see Figure 2.4). If H/L<< 0.1 then ks << H.
• Combining (2.22) with (2.23) one finds for a flat bed with D90 = 10 mm an n-value of 0.022. A
complete list of n-values for different bed conditions is given in Chapter 6.
Figure 2.4:
Hydraulic roughness as function of
dune/ripple dimensions
0.001 0.01 0.1 1
H/L
Dune/ripple dimensions
1.2
1.1
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
ks/H
For a dune/ripple covered bed the equivalent sand roughness ks and hence also Manning’s n-value
are not constant but will vary with flow depth and excess shear stress. Van Rijn (1984) developed the
following relations between dune/ripple dimensions, grain size and the transport stage parameter T,
defined below. The relations are valid for 0 < T < 5:

H
h
H
L
where: D50 = median grain diameter
Based on a large number of field data Julien and Klaassen (1995) found that for T > 5 relative dune
height and dune steepness are no longer a function of T. The following relations apply for T > 5:
H
h
= 
c

= 
c
c
D
50
h
D
H,1

H/ L,1
cH,1 = coefficient in dune/ripple height relation
cH/L,1 = coefficient in dune/ripple steepness relation
T = van Rijn’s transport stage parameter.
D
h
0.5T

−
≈ − − 
(1 e )(25 T) with :c 0.11
0.5T
H,1

−
≈ − − 
(1 e )(25 T) with :c 0.015
H/ L,1
< < =
with c and c
H
L
c

 
0.3
50
h

 
0.3
0 3
.
: . : .
50
0 8 8 25
2 2
H H H
D
h
, ,
< < =
with c and c

 

 
,
0 3
.
: . : .
2
2
50
012 2 0 4
2 2
H L H L H L
=
=
/ ,
/ , / ,
Substitution of (2.27) and (2.28) in (2.24) leads to:
0.3 and β = 10 D50
k s h { h } = α 0.7 1− exp(−β −0.3 )
For T > 5 the coefficients become: α = 3 D50
direction, this would mean that the equivalent sand roughness also decreases towards the river
mouth. Experience shows that this is not always the case.
The above equations provide a procedure to estimate the value of the hydraulic roughness based on
measurable and predictable quantities: bed-material size and dune/ripple dimensions. It can also be
used for design conditions, since it allows for extrapolation. In such cases it is necessary to calibrate
the dune-dimension relationship and roughness on local data in view of the large variation in the
coefficients cH and cH/L . To be able to carry out the computations the T-parameter has to be
determined.
Transport stage parameter T
The transport stage parameter T is a measure for the excess shear on the grains (shear stress above
the critical shear stress, where the latter indicates the initiation of motion) and is defined by:
θ '
−θ
θ
T cr
cr
=

2
12
where:
v
R
D g
θ' = : = log
 

 
2
Δ 3
C D
with C
g
18
50 90
θcr = dimensionless critical shear stress according to Shields; the latter is a function of D∗, defined by:
D
Δ
ν
g
D with
4 x
10 5
20
−
t c
∗
= 
 

 
=
+
ν
2
/
1 3
50
:
(2.25)
(2.26)
(2.27)
(2.28)
(2.29)
0.3. Since D50 reduces in downstream
(2.30)
(2.31)
(2.32)

where: tc = temperature in oC. The relation between D∗ and θcr is presented in Table 2.2. Note that the
Shields curve refers to the situation that a large number of particles are put in motion.
D∗ -range θcr
D∗ ≤ 4 0.24 D∗
4 < D∗ ≤ 10 0.14 D∗
10 < D∗ ≤ 20 0.04 D∗
20< D∗ ≤ 150 0.013 D∗
D∗ > 150 0.055
Table 2.2 Shields curve as function of D∗
The quantity Δ is the relative density of sediment under water: Δ =(ρs-ρ)/ρ (=1.65)
Hydraulic roughness for compound channels
In the above it has been indicated that a clear relationship exists between bed features and the
hydraulic roughness, whether it is expressed by ks or Manning’s n. In view of this it will be obvious that
a combined value for ks or n for a compound channel does not make sense. The values have to be
determined/estimated for each segment separately to be of any value for rating curve extrapolation!
2.5 UNSTEADY FLOW
The propagation and attenuation of flood waves in river systems are described by the following partial
differential equations (see Volume 4, Reference Manual):
• Continuity equation:
0
∂
+
∂
A Q
=
∂
x
∂
t
0
2
n QQ
-0.64
-0.1
0.29
− + =
4 / 3
R A
S
∂
+
-1
h
∂
x
• Momentum equation:
∂
Q
∂
t
1
gA
Q
A
1
∂
∂
gA x
2
s
0
2

+  
s s
s

 
(2.33)
(2.34)
where: A = total cross-sectional area (conveying and storage areas)
B = total width of cross-section
As = cross-sectional area of conveying section
Bs = width of conveying cross-section
h = flow depth
S0 = bottom slope
n = Manning’s hydraulic roughness parameter
R = hydraulic radius of conveying cross-section
A definition sketch of the cross-section is shown in Figure 2.5. The above equations form the so-called
Saint-Venant equations.

B
Bs
As
A
Storage area h
Conveying area
It can be shown, that, if the Froude number is small, the first two terms in (2.34), which represent
convective and local acceleration effects, are negligible compared to ∂h/∂x. Then, the momentum
equation reduces to:
1/ 2
∂
h
1
1


Q  
 
= −
A R S 1
n
s x
0
1/ 2
0
2 / 3
S
∂
From (2.35) it is observed, that the bracketed term approaches 1 if ∂h/∂x << S0, and the momentum
equation reduces to:
1/ 2
0
5 / 3
1
1
Q = ≈
A R S
n
B h S
s s
1/ 2
0
2 / 3
n
The latter expression applies for wide rivers, where As ≈ Bsh and R ≈ h. Because of the condition
∂h/∂x << S0 it follows that for unsteady flow (2.36) is typically only suited for steep rivers.
The celerity of a flood wave is given by:
c
Flood wave celerity
= ≈ 1
dQ
dA B
dQ
dh
Differentiation of (2.36) with respect to h and substitution into (2.37) gives the following expression for
the celerity of a flood wave or kinematic wave:
c
B
B
≈ s v 5
3
Figure 2.5:
Definition sketch river
cross-section
(2.35)
(2.36)
(2.37)
(2.38)
From this it is observed that for a river without a flood plain, i.e. Bs = B, the flood wave moves faster
than the average flow velocity. If, however, B >> Bs, i.e. for a river with a wide flood plain, then the
flood wave will move slower than the average velocity in the main river. Hence, it is observed that the
flood wave celerity will change if the river flow changes from inbank to overbank.

From the continuity equation (2.33) and the momentum equation (2.35) an approximate expression for
the damping of a flood wave per unit distance can be derived:
dQ
dx
Flood wave damping
D
c
2
Q
t
∂
∂
≈ where : D
=
Q
BS atQ
max
max
3
2
2 0
For a sinusoidal wave with amplitude a0 and duration/period T the wave damping becomes:
dQ ≈ − s (2.39)
Equation (2.39) shows that the damping of a flood wave is large, if:
• Total width of river and flood plain is large compared to the river width
• Hydraulic roughness is large
• Slope of the riverbed is small
• The flood wave amplitude is large, and
• The duration of the flood wave is small.
Hence, the steeper the flood wave the stronger it attenuates.
Looped stage-discharge relation
From equation (2.35) it is observed that for sub-critical flow there is no unique relationship between
stage and discharge. Since ∂h/∂x < 0 for the rising stage and > 0 thereafter, it is seen that for equal
stages the discharge is larger when the flow is rising than when the flow is falling. Since the rate of
rise is generally larger than the rate of fall the actual stage-discharge relation will behave
asymmetrical about the steady state rating curve. Because the slope of the water table is difficult to
monitor, by making use of the continuity equation (2.33) and (2.37) ∂h/∂x is replaced by –1/c.∂h/∂t,
which is measurable from the hydrograph observed at one station. Equation (2.35) then evolves to the
so-called Jones-equation, which reads:
∂
h
0
2
a
= +
( )
2 2
∂
n B B
max /
1/ 3
4.3
h S
dx
= +
2
0
T
s ∂
t
1
S c
Q 1
h
∂
t
1
S c
1
1/ 2
0
5 / 3
B h S 1
n
Q
0
s
0
(2.40)
where: Qs = steady uniform flow.
The looped stage discharge relation is shown in Figure 2.6.

Steady state stage-discharge
rating curve
A
B
Rising stages
Falling stages
h
Deviations A and B effect of unsteady flow,
2.6 BACKWATER CURVES
Qs=f(h)
Q=f(h,∂h/∂ t)
Q
generally A > B
Downstream tributaries, deltas, coasts, reservoirs, lakes, structures and aquatic vegetation growth
can all cause variable backwater effects which can effect the stability and reliability of the stage-discharge
a site is upstream of a reservoir or some other downstream influence it is possible using one of the
following methods to obtain an initial estimate of the possible impact of backwater. A decision can
then be made whether the site under consideration would be better located further upstream. The
terms used in the method are illustrated in Figure 2.7 below.
To estimate the extent of backwater some simple procedures are introduced here for a rapid
assessment. For a more detailed treatise reference is made to Volume 4, Reference Manual,
Hydrometry.
To describe the backwater curve use is made of the Bélanger equation, which reads:
dh
dx
S
relationship. Such effects should be avoided, if possible, during the site selection process. If
3 3
−
− 0
h h
h h
n
3 3
c
=
Figure 2.6:
Effect of unsteady flow
on stage-discharge
relationship
Figure 2.7:
Definition sketch for backwater
effect
(2.41)

where: hn = normal, equilibrium or uniform flow depth, from (2.36)
hc = critical flow depth at the transition from sub-critical to super-critical flow (Fr = 1),
from (2.4)
• Equilibrium or normal flow depth hn:
3 / 5
nq


=
h  
 
n S
1/ 2
0
where: q = discharge per unit width = Q/Bs = vh
n = Manning’s hydraulic roughness parameter
S0 = bed slope of the river
• Critical flow depth hc:
For given q, S0 and n, hn and hc are known quantities. So, (2.41) is an ordinary differential equation in
h.
Approximation of backwater effect
Assuming a gradually varied flow M1 type profile and a wide rectangular cross-section, a first order
estimate of the extent of backwater is obtained from:
3S L

2 1/3

 
−
h

q
g c =
 

 
 
h h exp 0 x
Δ << Δ ≈ Δ
x 0 for : h h
2 0 n
−
h (1 Fr )
n


where: Δhx = backwater effect at x = Lx
Note that this estimate applies for Δh0 << hn. A crude order of magnitude for the distance over which
the backwater is felt, is obtained from:
L
h
S x
≈ n
0
(2.42)
(2.43)
(2.44)
Δh0 = initial set up of water level at x = 0
S0 = river bottom slope
Lx = distance
(2.45)
For a compound cross-section in which river (r) and floodplain (f) both convey part of the total
discharge, hn in (2.22) and (2.45) is to be replaced by hE:
2
3 / 2
+
B h B h


h f   f

3 / 2
r r
+
B h B h
r r f f
E
 

=
This equation holds well if the roughness in the river and the flood plain does not differ much.
estimation by Bresse function

For more accurate computations one can apply the Bresse function. For a wide river the set of
equations to solve read (Chow, 1959):
[ { }]
γ = −
( ) ( ) ( )
x 0 x 0
; 1 Fr
  
  
η +
h
n
= − η − η − γ ψ η − ψ η
η =
−
0

 
L
x
where :
S
Δ +
h h
x n
n

 
;
2
η + η +
η −
η =
x
and
ψ η =
Δ +
2 1
3
h h
1
0 n
h
n
arc cot g
3
( 1)
1
1
h
ln
6
( )
2
2
(2.46)
The function Ψ(η) for 1 ≤ η < 1.25 is presented in Table 2.3. An application is presented in Example
2.1.
η Ψ η Ψ η Ψ η Ψ η Ψ
1.000 ∞ 1.050 0.896 1.100 0.681 1.150 0.561 1.200 0.480
1.001 2.184 1.051 0.889 1.101 0.678 1.151 0.559 1.201 0.478
1.002 1.953 1.052 0.883 1.102 0.675 1.152 0.557 1.202 0.477
1.003 1.818 1.053 0.877 1.103 0.672 1.153 0.555 1.203 0.476
1.004 1.723 1.054 0.871 1.104 0.669 1.154 0.553 1.204 0.474
1.005 1.649 1.055 0.866 1.105 0.666 1.155 0.551 1.205 0.473
1.006 1.588 1.056 0.860 1.106 0.663 1.156 0.550 1.206 0.472
1.007 1.537 1.057 0.854 1.107 0.660 1.157 0.548 1.207 0.470
1.008 1.493 1.058 0.849 1.108 0.657 1.158 0.546 1.208 0.469
1.009 1.454 1.059 0.843 1.109 0.655 1.159 0.544 1.209 0.468
1.010 1.419 1.060 0.838 1.110 0.652 1.160 0.542 1.210 0.466
1.011 1.388 1.061 0.833 1.111 0.649 1.161 0.541 1.211 0.465
1.012 1.359 1.062 0.828 1.112 0.647 1.162 0.539 1.212 0.464
1.013 1.333 1.063 0.823 1.113 0.644 1.163 0.537 1.213 0.463
1.014 1.308 1.064 0.818 1.114 0.641 1.164 0.535 1.214 0.461
1.015 1.286 1.065 0.813 1.115 0.639 1.165 0.534 1.215 0.460
1.016 1.264 1.066 0.808 1.116 0.636 1.166 0.532 1.216 0.459
1.017 1.245 1.067 0.804 1.117 0.634 1.167 0.530 1.217 0.458
1.018 1.226 1.068 0.799 1.118 0.631 1.168 0.528 1.218 0.456
1.019 1.208 1.069 0.795 1.119 0.628 1.169 0.527 1.219 0.455
1.020 1.191 1.070 0.790 1.120 0.626 1.170 0.525 1.220 0.454
1.021 1.175 1.071 0.786 1.121 0.624 1.171 0.523 1.221 0.453
1.022 1.160 1.072 0.781 1.122 0.621 1.172 0.522 1.222 0.451
1.023 1.146 1.073 0.777 1.123 0.619 1.173 0.520 1.223 0.450
1.024 1.132 1.074 0.773 1.124 0.616 1.174 0.519 1.224 0.449
1.025 1.119 1.075 0.769 1.125 0.614 1.175 0.517 1.225 0.448
1.026 1.106 1.076 0.765 1.126 0.612 1.176 0.515 1.226 0.447
1.027 1.094 1.077 0.760 1.127 0.609 1.177 0.514 1.227 0.445
1.028 1.082 1.078 0.756 1.128 0.607 1.178 0.512 1.228 0.444
1.029 1.071 1.079 0.753 1.129 0.605 1.179 0.511 1.229 0.443
1.030 1.060 1.080 0.749 1.130 0.602 1.180 0.509 1.230 0.442
1.031 1.049 1.081 0.745 1.131 0.600 1.181 0.507 1.231 0.441
1.032 1.039 1.082 0.741 1.132 0.598 1.182 0.506 1.232 0.440
1.033 1.029 1.083 0.737 1.133 0.596 1.183 0.504 1.233 0.438
1.034 1.019 1.084 0.734 1.134 0.594 1.184 0.503 1.234 0.437
1.035 1.010 1.085 0.730 1.135 0.591 1.185 0.501 1.235 0.436
1.036 1.001 1.086 0.726 1.136 0.589 1.186 0.500 1.236 0.435
1.037 0.992 1.087 0.723 1.137 0.587 1.187 0.498 1.237 0.434
1.038 0.983 1.088 0.719 1.138 0.585 1.188 0.497 1.238 0.433
1.039 0.975 1.089 0.716 1.139 0.583 1.189 0.495 1.239 0.432
1.040 0.967 1.090 0.713 1.140 0.581 1.190 0.494 1.240 0.431
1.041 0.959 1.091 0.709 1.141 0.579 1.191 0.492 1.241 0.429
1.042 0.951 1.092 0.706 1.142 0.577 1.192 0.491 1.242 0.428
1.043 0.944 1.093 0.703 1.143 0.575 1.193 0.490 1.243 0.427
1.044 0.936 1.094 0.699 1.144 0.573 1.194 0.488 1.244 0.426
1.045 0.929 1.095 0.696 1.145 0.571 1.195 0.487 1.245 0.425
1.046 0.922 1.096 0.693 1.146 0.569 1.196 0.485 1.246 0.424
1.047 0.915 1.097 0.690 1.147 0.567 1.197 0.484 1.247 0.423
1.048 0.909 1.098 0.687 1.148 0.565 1.198 0.483 1.248 0.422
1.049 0.902 1.099 0.684 1.149 0.563 1.199 0.481 1.249 0.421
Table 2.3: The function Ψ(η) for 1 ≤ η < 1.25

Example 2.1 Application of Bresse function
The distance is to be determined over which the backwater effect at x = 0 of 1 m is reduced to 5% of its
original value for a river with a bed slope of 5x10-4, hydraulic roughness n = 0.03, when the normal depth hn
= 5 m.
1. Determination of η and Ψ(η)
For x = 0 the backwater is Δh0 = 1m hence η0 becomes:
1.20
+
1.0 5.0
Δ +
η =
0 =
5.0
h h
0 n
h
n
=
If x = x1 is the distance at which the initial backwater effect is reduced to 5% of its value: Δh1 = 0.05 m,
hence η1 becomes:
1.01
+
0.05 5.0
Δ +
η =
1 =
5.0
h h
1 n
h
n
=
It then follows for Ψ(η0) and Ψ(η1) from Table 2.3: Ψ(η0) = 0.480 and Ψ(η1) = 1.419
2. Froude correction γ:
−
1/3 4
5 x5x10
γ 2 = − = − = − =
9.81x0.03
1
1/3
n
h S
gn
1
v
2
gh
= 1- Fr 1
2
2
0
The parameter γ in equation (2.46) follows from:
3. Computation of Lx
0.903
The distance Lx = x1 – x0 follows from (2.46) by substitution of the values determined under 1 and 2:
h
h
[ { }] [ ] n
10,380m 10.4km
x = − η − η − γ Ψ η − Ψ η = − − − − = = =
S
(1.01 1.2) 0.903x(1.419 0.480) 1.038x
n
S
( ) ( ) ( )
h
n
S
L
0
0
1 0 1 0
0
Note that the distance is only 4% larger than one would have obtained from (2.45). The results (Lx at 5% of
the original value, expressed as a function of hn/S0) for different river slopes and roughness values for the
same normal depth (5 m) and initial backwater (1 m) are presented in the following table:
Froude parameter γ Lx expressed as function of hn/S0
Roughness n Roughness n
S0
0.025 0.03 0.05 0.025 0.03 0.05
1x10-3
5x10-4
1x10-4
0.721
0.861
0.972
0.806
0.903
0.981
0.930
0.965
0.993
0.867
0.998
1.103
0.947
1.038
1.111
1.063
1.096
1.122
It is observed, that the multiplier to hn/S0, to arrive at Lx , is close to 1 for different river slopes and
roughness values. Adding some 10% to the value for Lx obtained from (2.45) will give a reasonable
approximation of the extent of the backwater reach in practice for field applications.

3 HYDROMETRIC NETWORK DESIGN
3.1 INTRODUCTION
A hydrometric network is a system of river gauging stations in a river basin at which river stage and
discharge are measured. The network provides hydrologic data needed for the planning, design and
management of conservation and utilisation of the waters and other natural resources of the river
system. In flood prone areas the network may also provide data for design and management of flood
protection measures. The data should enable accurate estimation of the relevant characteristics of the
hydrological regime of the river basin. The network requirement is greatly influenced by a number of
factors including:
• monitoring objectives, determined by the data needs of the hydrological data users
• temporal and spatial variability of the river flow, determined by:
• climatic features like precipitation pattern in the catchment, evapo(transpi)ration
• physiographic features of the river basin, like size, slope, shape, soils, land use and drainage
characteristics
• the availability of financial, manpower and other resources.
The identification of the monitoring objectives is the first step in the design and optimisation of the
monitoring systems. Related to this is the identification of the potential data users and their future
needs. Reference is made to Chapter 3 of Volume 1 of the Design Manual on Hydrological
Information System for a summary. The actual data need for a particular basin is to be obtained by
interviewing the potential hydrological data users, to be presented in a Hydrological Information Need
(HIN) document, where in case of more objectives, priorities are indicated.
The second variable to be considered in the design of the hydrometric network is the dynamics of the
river flow and stages in time and space. This requires a critical analysis of historical data.
To enable an optimal design of the monitoring system a measure is required, which quantifies the
effectiveness level. This measure depends on the monitoring objectives and can be related to an
admissible error in e.g. the mean flow during a certain period, monthly flow values for water balances,
extreme flows and/or river stages, etc. This error is a function of the sampling locations, sampling
frequency and sampling accuracy, i.e. where, when and with what are river/reservoir stages and flows
to be measured.
Reference is made to Chapter 7 of Volume 2 of the Design Manual on Sampling Principles for an
introduction into the general principles of network design and optimisation. In this volume the
principles are tuned to the hydrometric network. It is, however, stressed that the hydrometric network
should never be considered in isolation. The network is part of an integrated system of networks of
the HIS including also hydro-meteorology, geo-hydrology and water quality. The totality of the
networks should provide the data requested for by the Hydrological Data Users.
In this chapter the following topics are discussed:
• general hydrometric network design considerations,
• network density, and
• network design process.

3.2 NETWORK DESIGN CONSIDERATIONS
In this section a number of aspects are discussed to be considered before actually designing the
hydrometric network, including:
• Classification of stations,
• Minimum networks,
• Networks for large river basins,
• Networks for small river basins,
• Networks for deltas and coastal flood plains,
• Representative basins,
• Sustainability,
• Duplication avoidance, and
• Periodic re-evaluation.
3.2.1 CLASSIFICATION
Based on the network levels presented in Sub-section 7.2 of Volume 2, Design Manual, Sampling
Principles the following classification of stations is introduced:
Primary stations, maintained as key stations, principal stations or bench mark stations, where
measurements are continued for a long period of time to generate representative flow series of the
river system and provide general coverage of a region.
Secondary stations, which are essentially short duration stations intended to be operated only for
such a length of period, which is sufficient to establish the flow characteristics of the river or stream,
relative to those of a basin gauged by a primary station.
Special purpose stations, usually required for the planning and design of projects or special
investigations and are discontinued when their purpose is served. The purpose could vary from
design, management and operation of the project to monitoring and fulfilment of legal agreements
between co-basin states. The primary as well as secondary stations may also, in time serve as
special purpose stations.
In designing a network all types of stations must be considered simultaneously.
3.2.2 MINIMUM NETWORKS
A minimum network should include at least one primary streamflow station in each climatological and
physiographic area in a State. A river or stream, which flows through more than one State, should be
gauged at the State boundary. At least one primary gauging station should also be established in
those basins with potential for future development.
A minimum network should also include special stations. Where a project is of particular socio-economic
importance to a State or Region it is essential that a gauging station is established for
planning, design and possibly operational purposes. Sometimes special stations are required to fulfil a
legal requirement e.g. the quantification of compensation releases or abstraction controls. Benefit -
cost ratios for special stations are usually the highest and can help support the remainder of the
hydrometric network.

3.2.3 NETWORKS FOR LARGE RIVER BASINS
A primary station might be planned at a point on the main river where the mean discharge attains its
maximum value. For rivers flowing across the plains, this site is usually in the downstream part of the
river, immediately upstream of the point where the river normally divides itself into branches before
joining the sea or a lake or crosses a State boundary. In the case of mountainous rivers, it is the point
where water leaves the mountainous reach and enters the plain land. Subsequent stations are
established at sites where significant changes in the volume of flow are noticed viz., below the
confluence of a major tributary or at the outflow point of a lake etc.
If a suitable location is not available below a confluence, the sites can be located above the
confluence, preferably on the tributary. While establishing sites downstream of a confluence, care
should be taken to ensure that no other small stream joins the main river so as to avoid erroneous
assessment of the contribution of the tributary to the main river. In the case of a large river originating
in mountains, though the major contribution is from upper regions of the basin, several stations may
have to be located in the downstream stretch of the river. Such stations are intended to provide an
inventory of water loss from the channel by way of evaporation, infiltration, and by way of utilisation for
irrigation, power generation, industrial and other domestic needs.
The distance between two stations on the same river may vary from thirty to several hundred
kilometres, depending on the volume of flow. The drainage areas computed from origin up to
consecutive observation sites on a large river should preferably differ by more than 10% so that
the difference in quantities of flow is significant. The uncertainties in discharge values particularly
for high flows are unlikely to be less than +/- 10%. However, every reasonable attempt should be
made to minimise these uncertainties.
The above uncertainties may affect the location of stations. When tributary inflow is to be known it is
generally better to gauge it directly, rather than deriving the flow from the difference of a downstream
and an upstream station along the main stream (see Volume 2, Design Manual, Chapter 4, Example
4.1). Also, a more accurate discharge record for the main stream is obtained from monitoring the
feeder rivers than by a main stream station alone, however, at the expense of additional cost.
3.2.4 NETWORKS FOR SMALL RIVER BASINS
The criteria mentioned in Sub-section 3.2.3 are applicable to a river basin having a large area and
well developed stream system. A different approach is to be adopted in dealing with small
independent rivers, which flow directly into the sea, as in the case of west flowing rivers of Kerala and
Maharashtra and some east flowing rivers of Tamil Nadu. In such cases, the first hydrological
observation station might be established on a stream that is typical of the region and then further
stations could be added to the network so as to widely cover the area. Streams in a particular area
having meagre or lower yields should not be avoided for inclusion in the network. Absence of a station
on a low flow stream may lead to wrong conclusions on the water potential of the area as a whole,
evaluated on the basis of the flow in the high flow streams. Thus, great care is to be exercised in
designing the network to ensure that all distinct hydrologic areas are adequately covered. It is not
possible to operate and maintain gauging stations on all the smaller watercourses in the Western
Ghats, for example. Therefore, representative basins have to be selected and the data from those are
used to develop techniques for estimating flows for similar ungauged sites.
3.2.5 NETWORKS FOR DELTAS AND COASTAL FLOODPLAINS
Deltaic areas such as the Lower Mahanadi in Orissa, where gradients are usually low and channels
bifurcate, are often important, as water use is productive and thus these areas need monitoring. This
is particularly important, as deltas are dynamic systems, i.e. they are continually changing. However,
the type of network required may differ from more conventional river basins. It is often not possible
due to the low gradients to locate stations with stable stage-discharge relationships, i.e. variable
backwater effects can occur due to tidal influences and/or changes in aquatic vegetation growth.

Stage readings should be made at all principal off-takes/bifurcations or nodes in the system. These
should be supplemented by current meter gaugings when required. At some sites consideration might
be given to installing a slope-area method station.
3.2.6 REPRESENTATIVE BASINS
When gauging stations are included in a network to obtain representative data from a particular
physiographic zone, it is better if the chosen basins are those with the water resource relatively under
utilised, i.e. the basins can be considered to be close to their natural state. The selection of
representative gauging stations in basins, which are heavily utilised by dams and water abstraction
and/or where significant land use changes have and are continuing to take place should be avoided.
3.2.7 SUSTAINABILITY
Of paramount importance is sustainability. It is a relatively straightforward task to design a dense
network of streamflow stations. However, the implementation and operation of a network is a lot more
difficult. It has been found from experience, that there is a tendency to adopt an idealistic approach
and attempt to have as many stations as possible. There are many examples of networks throughout
the world, which are no longer functioning well due to lack of financial support, skilled manpower and
logistic support resources such as vehicles. It is far better to operate and maintain 10 gauging stations
well than to operate and maintain 20 stations badly i.e. higher quality data from fewer stations is
preferable to a lower quality of data from a greater number of stations.
3.2.8 DUPLICATION AVOIDANCE
Since, generally more than one organisation is responsible for the establishment of gauging stations
e.g. the State Water Departments and CWC, it is essential that the activities are co-ordinated so
they complement each other and duplication of effort is avoided.
3.2.9 PERIODIC RE-EVALUATION
Gauging station networks require periodic re-evaluation. The developments that take place in the
basin, like construction of new irrigation/hydro-electric projects and industrialisation of the area, may
warrant addition or closure of stations. For example river reaches are often polluted due to the
discharge of effluents from industry. Therefore, a need may arise to establish stations to assist with
water quality monitoring and pollution assessments.
3.3 NETWORK DENSITY
The World Meteorological Organisation developed guidelines on minimum hydrological network
densities. Their guidelines and potential use and limitations are presented in this section.
Furthermore, a prioritisation system is introduced to rank the importance of stations. Finally,
comments are given on the use of statistical and mathematical optimisation techniques for
hydrometric networks.
3.3.1 WMO RECOMMENDATIONS
The recommendations of the WMO (World Meteorological Organisation) on the minimum density of a
streamflow network for regions with different physiographic features are reproduced in Table 3.1
below.

Type of region Range of norms for
minimum network
Range of provisional norms
tolerated in1 difficult
conditions
I. Flat regions 1,000 - 2,000 3,000 - 10,000
II. Mountainous regions 300 - 1,0002 1,000 - 5,000
III. Arid zones3 5,0004 - 20,000 ------------------
NOTES:
1. Last figure in the range should be tolerated only for exceptionally difficult conditions;
2. Under very difficult conditions this may be extended up to 10000 km2;
3. Great deserts are not included;
4. Under very difficult conditions this may be extended up to 10000 km2.
Table 3.1: Minimum density of hydrological network according to WMO, area in km2 for one
station
It is not possible to provide specific, general guidelines on an appropriate network density. The WMO
recommendations are very general guidelines which if adopted at face value for some of India’s larger
river basins could result in an excessively dense network. Even though the WMO type guidelines
might be used as rough rule of thumb as part of an initial network appraisal, their use in the
final design of the network should be avoided. The network density must ultimately be based on
the network objectives, the temporal and spatial variability of river stages and flow and on the
availability of finance, manpower and other resources.
3.3.2 PRIORITISATION SYSTEM
It is suggested that in the first instance the “ideal” network size is determined. In determining the
network all potential users of the data should be consulted. Each station in the “ideal” network should
be prioritised. In order to do this a simple prioritisation system is useful. This prioritisation system
could be a simple one such as follows:
Category Priority Relative Importance
A High Major, multi-purpose water resources development site, State boundary river, operation of
major scheme, major ungauged basin, heavily polluted major water supply source
B Medium Medium scale water resources development project site, secondary basin, industrial
development area, i.e. potential water quality problems)
C Low Minor irrigation project site, secondary gauging station on tertiary tributary, major water
course but already extensively gauged
The above categories and priorities are merely highlighted by way of example. Each State/Central
organisation needs to set its own priorities based on its own policies and objectives. In prioritising
sites, the following questions should be asked:
What are the socio-economic consequences of not collecting streamflow data at the site?
What are the alternatives to establishing a streamflow gauging station at the site under consideration?
An estimate of the number of stations within each State, Division and Sub-division which can be
realistically well maintained should be made. When deriving this estimate, the following factors should
be considered:
• The recurrent budget implications;
• Short and longer term manpower requirements and availability of suitably skilled personnel;
• Capacity of instrument repair, spare part provision and calibration facilities;
• Long term availability of logistic support facilities such as vehicles.

The ideal and realistic network size estimates should be compared. If necessary the size of the ideal
network should then be reduced by removing the lower priority stations.
3.3.3 STATISTICAL AND MATHEMATICAL OPTIMISATION
The streamflow network should provide information for the location indicated by the hydrological data
users. At a number of locations no stations will be available. Hence the information is to be obtained
from the network by e.g. interpolation. If the interpolation error in estimating a flow characteristic is too
large than additional stations or a re-design should be considered. These techniques are most
applicable to already well established networks, where the data have been rigorously quality
controlled and are readily available in computer compatible form. However, they are less readily
applied to heavily utilised, over-regulated catchments like many of the larger river basins in India.
These techniques are a tool to assist with network design. They are not straightforward to apply and
do not totally obviate the need for the pragmatic, common sense approach.
3.4 THE NETWORK DESIGN PROCESS
Since everywhere hydrometric networks are existing, the network design process is one of evaluation,
reviewing and updating of an existing network. The historic evolution of many hydrometric networks
has tended to be reactively rather than strategically planned. Often gauging stations are being
operated for which the original objectives are unclear. It is therefore necessary to regularly undertake
a detailed review of the existing networks to achieve the following:
• Define and/or re-define the purpose of each gauging station;
• Identify gaps in the existing network;
• Identify stations which are no longer required;
• Establish a framework for the continual evaluation and updating of the network.
There is a tendency for hydrometrists, hydrologists and water resources planners to be reluctant to
discontinue gauging stations, even though they might have fulfilled their intended objectives. In the
design and evaluation of networks it is essential that a ‘hard nosed’ approach is adopted and
stations which are no longer providing a significant benefit are discontinued.
In Chapter 7 of Volume 2, Design Manual, Sampling Principles a list of steps is presented to be
carried out for the review and redesign design of a network. Specific steps for the review of the
hydrometric network are outlined in Part I of Volume 4, Field Manual, Hydrometry. The main steps in
the network design process can be summarised as follows:
1. Review mandates, roles and aims of the organisations involved in the operation of the HIS in a
particular area and evaluate the communication links.
2. Collect maps and other background information.
3. Define the purposes of the network: - who are the data users and what will the data be used for?
Define the objectives of the network: - what type of data is required where and at what
frequency?
4. Evaluate the existing network: - How well does the existing network meet the overall objectives?
5. Review existing data to identify gaps, ascertain catchment behaviour and variability.
6. Identify gaps and over-design in the existing network: Propose new stations and delete existing
stations where necessary, i.e. revise the network.
7. Prioritise gauging stations: i.e. try to use some simple form of classification system.

8. Estimate average capital and recurrent costs of installing and maintaining different categories of
hydrometric stations. Estimate overall cost of operating and maintaining the network.
9. Review the revised network in relation to overall objectives, ideal network, available budgets and
the overall benefits of the data. Investigate the sustainability of the proposed network.
10. Prepare a phased implementation plan. This has to be prioritised, realistic and achievable.
11. Decide on the approximate location of sites, commence site surveys. If a site is not available
review the location and see if another strategy can be adopted, e.g. gauge a tributary to estimate
total flow at the required spot rather than trying to measure the total flow in the main stem river.
Guidelines on site selection are contained in Chapter 4.
12. Establish a framework for regular periodic network reviews.
As hydrometric network design is a dynamic process, networks have to be continually reviewed and
updated so that they react to new priorities, changes in policies and fiscal changes. Regular
formalised network reviews should be undertaken, recommended to take place after 3 years or at a
shorter interval if new data needs do develop.

4 SITE SELECTION OF WATER LEVEL AND STREAMFLOW
STATIONS
4.1 DEFINITION OF OBJECTIVES
Prior to embarking on the selection of a water level (stage) monitoring or streamflow-gauging site, it is
imperative that the objectives of the site are fully defined. In this regard the following factors have to
be established:
1. What is the purpose of the station? E.g. planning and design of major water supply scheme,
pollution monitoring, flood forecasting, etc.
2. Define the required location of the site, i.e. what are the most upstream and downstream limits;
e.g. the station might have to be located between two major tributaries.
3. Does the full flow range require monitoring or are low or high flows of greater importance?
4. What level of accuracy is required?
5. What period of record is required and what frequency of measurement is desirable?
6. Who will be the beneficiaries of the data?
7. Is a particular streamflow measurement methodology or instrument preferred?
8. Are there any constraints such as access and land acquisition problems and cost limitations?
The activities involved in selection of a site are dealt with in Section 4.3. The selection of water level
gauging and discharge measuring sites are discussed in respectively Sections 4.4 and 4.5. The
selection of a streamflow measuring site requires a proper understanding of the various types of
controls; an overview is presented in Section 4.2.
4.2 DEFINITION OF CONTROLS
Terminology
The shape, reliability and stability of the stage-discharge relationship are normally controlled by a
section or reach of channel at, or downstream of the gauging station, which is known as a control. In
terms of open channel hydraulics a control is generally termed a section control if critical flow occurs
a short distance downstream from the gauging station. This can occur where a natural constriction or
a downward break in channel slope occurs resulting from a rock outcrop or a local constriction in
width caused by the construction of a bridge. If the stage-discharge relationship depends mainly on
channel irregularities and friction downstream of the station then this is referred to as a channel
control. This is the most common type of control in India. A complete control is one which
determines the stage-discharge relationship throughout the complete range of flow e.g. at a high
waterfall. However, more commonly no single control is effective for the entire range and we then
have a compound control. This could be a combination of a section control at low stages and
channel control for high stages. A control is permanent if the stage-discharge relationship it defines
does not change with time, otherwise it is referred to as a shifting control.
Controls can ether be natural or artificial (man made for flow measurement purposes). Artificial
controls may be purpose built flow measurement structures, which have a theoretical stage -
discharge relationship unique to the structure. As such it is not necessary to undertake a large
number of current meter gaugings in order to define the stage - discharge relationship. However, the

theoretical relationships should be checked over the full range of flows. Reservoir spillways, control
weirs and anicuts frequently come into the ‘artificial’ category, even though they have not been
purpose built for flow measurement purposes, since it is often possible to derive theoretical stage-discharge
relationships. Structures, which have not been constructed for the purpose of flow
measurement such as bridges, floodway channels and drifts, are not considered as artificial controls
since they normally require full calibration.
Stage - discharge gauging stations such as natural controls and non-purpose built structures, which
require current meter gauging to define the stage-discharge relationship are often referred to as rated
sections.
The two most important attributes of a control are stability and sensitivity (the two “S”s). If the
control is stable the stage-discharge relation will be stable. It is also important that the control is
sensitive, i.e. small changes in water level should correspond to relatively small changes in discharge.
Hydrometric sensitivity
It is a primary requirement for stage-discharge gauging stations that the rating relationship should be
as sensitive over as wide a range of flows as possible. In other words, any change in the recorded
water level should correspond to a relatively limited (in percentage rather than absolute terms) change
in flow. This is illustrated by the rating curves sketched below, see Figure 4.1. From this sketch it can
be seen, that rating 1 is more sensitive than rating 2.
Figure 4.1:
Sketch illustrating the concept of
hydrometric sensitivity
h
Δh
ΔQ1 ΔQ2
4.3 SITE SURVEYS
Rating 1
Rating 2
Q
ΔQ1<< ΔQ2 for same Δh
Once the purpose and objectives have been defined and the engineer responsible has considered
what flow measurement and automatic water level recording techniques could be suitable, the site
selection process can begin. The final choice of site will depend on the type and quality of the data
required, the method to be deployed and other factors such as logistics and budgetary constraints. In
particular the final site selection might be mainly determined by the choice of the most appropriate
equipment or technique. Therefore, some guidance is provided in Chapter 6, on the advantages and
limitations of different hydrometric methods used in, or which are suitable for Indian conditions.
In order to select the most appropriate site, considerable effort needs to be expended undertaking site
selection surveys. The site selection surveys can be divided into four distinct phases, which are
summarised in the sub-sections below:

1. Desk study
2. Reconnaissance surveys
3. Topographic surveys, and
4. Other survey work
To ensure that all the pertinent information is obtained during the site selection process and surveys
and to assist with the work, a standard form has been prepared. This list has been derived from the
CWC checklist. A copy of this form is contained in Appendix 2.1 of Chapter 2, Volume 4, Field
Manual, Hydrometry.
Desk study
The target location for the gauging station will have already been identified on a 1:250,000 map or
similar during the network design process. However, this size of map is too small a scale for site
selection purposes. The inspection of large-scale topographic maps (1:50,000) and aerial
photographs, if these are available, should be undertaken to identify possible sites within the target
river reach.
Reconnaissance surveys
These should be undertaken by road, foot and for larger, navigable rivers by boat. It is important that
the entire target reach of the river is inspected. During the survey, interviews should be held with local
people to try and build up a picture of the local site conditions such as water level ranges. At sites of
interest attempts could be made to ascertain who owns the land.
Topographic surveys
On completion of the reconnaissance surveys, one or more sites could have been identified which are
worthy of further consideration. However, it is often not possible to make final decisions on site
selection without the benefit of bed surveys. Cross-sectional surveys upstream and downstream of
the gauging site have to be carried out to get a detailed picture of the approach conditions and of the
layout and extent of the control section. Longitudinal profiles are to be obtained to assess the bed
slope and to identify potential controls. Possible backwater sources have to be identified.
Other survey work
It is useful to carry out flow measurements and stage observations at locations near to the proposed
site(s) prior to making the final site selection. From such measurements some idea will be obtained
about the velocity distribution across the measuring cross-section, water level variation and water
surface slope. If structures are to be installed it is possible that soils and geological surveys will be
required to establish the stability of the banks and bed for founding the structure and the availability of
construction material such as rip-rap.
Summing up
Once the surveys have been carried the following aspects have to be reviewed:
• technical aspects: the hydraulic suitability of the site,
• logistical aspects: site accessibility, communication and staffing,
• security aspects: security of instruments, away from residential areas and play grounds,
• legal aspects: land acquisition and right of passage, and
• financial aspects, including costs of land acquisition, civil works, equipment, data processing, and
staffing and training.

4.4 SELECTION OF WATER LEVEL GAUGING SITES
Stage measurements are most commonly required in surface water hydrometry to determine the flow
using relationships between stage and discharge or cross-sectional area and velocity. Therefore, in
many circumstances the selection of a stage or water level measurement site will be to a certain
extent governed by the suitability of the site for flow measurement purposes. It is very important that
extreme care is given to the selection of the location of stage monitoring devices since this is the
basic raw data, which is required to derive discharge. Also, in some situations, flow estimates will not
be required but it is necessary to measure water level only, e.g. in reservoirs, for flood warning.
The gauging site should be located outside high turbulence zones, close to the edge of the stream at
a place where the banks are stable and preferably steep. The downstream control shall be stable and
sensitive to be able to establish a stable stage-discharge relation, where significant changes in the
discharge create significant changes in stage. The site shall be outside the backwater zone of
confluences and structures. The extent of the backwater reach L is approximately (see Chapter 2):
(4.1)
h
L ≈ n
S
where: L = approximate reach of the backwater effect
hn = normal or equilibrium depth (to be replaced by hE for a compound cross-section)
S = slope of the river bed
Nearby benchmarks should be available or be established to allow regular levelling of the gauge.
Detailed selection criteria for water level gauging sites are presented in Chapter 2 of Part I of Volume
4, Field Manual on Hydrometry.
4.5 SELECTION OF STREAMFLOW MEASUREMENT SITE
The majority of streamflow measurement techniques are based on the velocity area method. Even
though the use of float measurements is sometimes inescapable, current meter gauging is the most
widely favoured velocity-area method technique. For most situations the same general site selection
criteria can be applied to each technique. The current Indian Standards on velocity-area method site
selection (see Reference Manual) and current international practice (e.g. ISO 748) have been
reviewed along with other considerations and a recommended set of guidelines have been prepared
for
• for current meter gauging sites
• for float measurement
• for discharge monitoring by Acoustic Doppler Current Profiler (ADCP)
• for Slope-area Method
• for selection of Natural Control (rated section) station site, and
• for selection of Artificial Control Sites.
For a stage-discharge station, both a stage measurement device and a current meter gauging site are
required in the same locality. However, it might not always be appropriate to locate the current
meter gauging site immediately adjacent to the stage measurement device since some of their
site selection criteria are different.
Detailed sets of guidelines for the distinguished measurement techniques are presented in Chapter 2
of Part I of Volume 4, Field Manual on Hydrometry, which should be carefully considered. Some
important criteria are presented below.

Current meter gauging site
The selected site shall have a long, straight, uniform, well-defined approach channel upstream of the
measuring section to ensure parallel and non-turbulent flow and to minimise irregular velocity
distribution. In practice, the approach length is related to the channel width. Generally, for rivers less
than 100 m wide a straight approach of 4 x channel width is considered to be sufficient, whereas for
rivers greater than 100 m wide the current Indian minimum standard of 400 m straight approach
should be adopted if possible. When the length of the straight channel is restricted it is recommended
that the straight length upstream should be at least twice that downstream. The site shall be year
round accessible and the section be stable, confined to one channel with no overbank flow. Sufficient
flow depth should be available to provide effective immersion of the current meter and the flow
velocities shall be within the calibration range of current meters (> 0.15 m/s and < 3.5 m/s). Above
criteria also apply for other sites, with some additions.
Float measurement sites
Float measurements require a measuring track, which is straight and uniform in cross-section over a
length of five times the average width of cross-sections. The riverbank shall be easily accessible to
mark the passage of the floats and wind effects shall be minimum.
Acoustic Doppler Current Profiler (ADCP) sites
The ADCP is a device for measuring velocity, direction and cross-section. As such it is a velocity area
device. However, in view of its technology it can cope with irregular velocity distributions and skew
flow conditions. The choice of measuring cross-section is therefore not so critical as other velocity-area
methods. The cross-section should, however, be free of rock or other objects to avoid damage to
the face of the transducers. The equipment requires at least 1.5 m water depth below the transducers,
which in turn are at least 0.3 m below the water surface. Hence, deep river sections are preferable.
For safety reasons average velocities should not exceed 4 m/s, since the instrument will be boat
mounted.
Slope-area method discharge estimation site
The site to which the slope-area method is applicable is straight and has uniform, stable cross-sections,
with uniform hydraulic characteristics. Sufficient fall in the water table shall be available in
the river reach to allow accurate determination of the water surface slope.
Natural control (rated section) station site
In practice there is very rarely an “ideal” location for a natural control (rated section) gauging station. It
is often required to compromise and to establish stations in far from ideal conditions. The site
selection is based on hydraulic criteria and on tactical considerations. The natural control should be
selected where the relationship between stage and discharge is substantially consistent and stable, not
affected by any significant backwater effect. The control shall be sensitive, such that a significant change
in discharge, even for the lowest discharges, should be accompanied by a significant change in stage.
Small errors in stage readings during calibration at a non-sensitive station can result in large errors in the
discharges indicated by the stage-discharge relationship. Attention is to be given to land acquisition,
security and staff availability.

Artificial control sites
There is a variety of different flow measurement structures. The choice of structure will depend on a
number of factors including objectives, flow range, afflux, size and nature of the channel, channel
slope and sediment load, operation and maintenance, passage of fish and not least, cost. The
applications and limitations of a structure will determine where its use is most appropriate. In this
regard each type of structure has its own specific site selection criteria. In addition to the above
mentioned criteria artificial controls also require appropriate sub-soil conditions to provide a solid
foundation for the structure. Since the structure will create some backwater effect, it may cause extra
flooding.

The frequency with which hydrological measurements are taken depends upon a number of factors,
such as:
1. The function which the data will serve. More frequent observation is required for the
assessment of peak flows especially for small catchments, than for some other purposes. Most
hydrological measurements are made to serve multiple functions. The measurement frequency
must meet the requirements of the uses planned for the data.
2. The target accuracy of derived data. It must be recognised that a policy of ‘as good as
possible’ may lead on occasions to unnecessary expenditure on improving accuracy beyond
what is needed for the purpose. i.e. there has to be a balance between the value of increased
accuracy of data and the increased cost of providing that increased accuracy. There must at least
be a notional upper limit on the size of the uncertainty band (accuracy) outside which the quality
of the data would generally be unacceptable for its intended uses. Conversely, this uncertainty
band should not be so small that the cost of providing data to such a high accuracy cannot be
justified by the end result.
The accuracy of derived hydrological data will depend on the sampling density, the accuracy of
measurement of the variable and the frequency of measurement. Sampling density determines
the representativeness and spatial variability of the derived data and is mainly covered in network
design (Chapter 3). The frequency of measurement as well as the accuracy of observation at a
specific location will have an impact on the accuracy of the derived data determination for that
point, e.g. mean daily flow, derived from a given number of stage values per day each of which is
converted to flow, (see para. 4 below).
3. The accuracy of observation. Where observations are subject to random measurement errors,
a larger number of observations are needed to meet a required target accuracy where the
measurement error is large. The standard error of the mean relationship (Smr) in a stage
discharge relationship is dependent both on the standard error (Se) of the observations and on the
number of observations (N), thus:
100
− 
Q Q

5 MEASURING FREQUENCY
5.1 GENERAL
= = Σ
 
g r
× mr Q
and : d
d
2
−
N 2
where : S
S
e
N
S
r
e

 
=
(5.1)
with: d = relative deviation of gauged flow Qg from the fitted stage-discharge relationship Qr
N = number of gaugings used to define the stage-discharge relation
4. The time variability of the variable. Fewer measurements are needed to determine the mean of
a variable over given time, if the variable is uniform or changing very slowly than for a rapidly
fluctuating variable. This is particularly important for the assessment of mean daily flow, a
common basis for many hydrological studies. Small steep upland catchments respond rapidly to
storm rainfall which itself fluctuates in intensity throughout a storm, giving a hydrograph, showing
rapid rise and fall and sharply defined peaks. On flat lowland basins the hydrograph is smoothed
by the variable timing of tributary inflows (though they may themselves be flashy) and by channel
and reservoir storages. Smaller catchments thus need more frequent measurement in order to
achieve the same accuracy of mean daily flow, than larger flat catchments and also for obtaining
a sufficient number of points to define the hydrograph of an event. Many hydrological variables
show a regular diurnal variation. This may either be natural as for climatic variables, arising from
the periodicity of solar energy input, or, man-made, for example daily cyclic changes due the
discharge of effluents or abstractions for irrigation. The frequency of measurements must be
sufficient to define the mean over both the highs and lows of these periodicities. Reference is
made to Chapter 4 of Volume 2, Design Manual, Sampling Principles, dealing with sampling at
the Nyquist frequency and beyond.

5. The seasonality of the variable. Flows in rivers or streams are highly seasonal and during the
monsoon the changes in stage and discharge can be rapid and large. Thus the frequency of
measurement of stage has to reflect this. Therefore, during the monsoon season a higher
measurement frequency is required (hourly or less). It can also be further reduced if an event is
taking place i.e. more manual readings or event trigger set in data logger. The past manual
practice in many States has been for daytime hourly observations of river level during the lean
season and 24-hour readings during the monsoon. The latter requires two additional staff (and the
associated increased costs) to ensure record continuity. Occasional out-of-season storms are
missed by daytime reading only. Where variables are observed and recorded autographically
(chart) and especially digitally, the required data for these out-of-season storms can be captured
at negligible increased handling and processing cost (see next para.)
6. The marginal cost/cost-benefit of improved accuracy. If manual observations are undertaken,
increasing the number of observations to improve the definition (accuracy) may require more
observer time (man-days). Therefore, costs will increase proportionally and decisions have to be
made on the benefits of the increased accuracy. However, where automated observations are
already taking place either by chart or DWLR, the frequency of observation can be increased with
only a negligible increase in cost.
7. The benefits of standardisation. It is simpler to process and analyse records which are arriving
at the Data Processing Centre, all in the same format and with the same frequency of
observation. This is true, both for manual data and for digital data, where batch processing of
records by computer is simplified. It may be preferable for digital observations to standardise on a
time interval close to the minimum requirement, than to adjust the frequency at each station to its
functions, accuracies and time variability of the parameter being measured, many of which are
imprecisely known.
In the subsequent sections the required sampling frequencies for stage and discharge are dealt with.
5.2 STAGE MEASUREMENT FREQUENCY
The procedure to determine the sampling frequency is outlined in Chapter 5 of Volume 2, Design
Manual, Sampling Principles. For full reproduction of the stage variation a sampling interval slightly
less than the Nyquist interval should be applied. This not only applies when interest is in peak flows,
but when rates of rise and of fall of the hydrograph have to be reproduced to accurately determine the
discharge of flashy floods in flat rivers. To estimate the mean flow over a certain period of time a
larger interval will suffice. Below, the sampling frequencies suggested for various types of rivers and
measurement techniques are presented. The frequencies given are indicative, based on past
experience. It is, however, necessary to verify the validity of these frequencies in view of the
objectives.
In larger flatter rivers of Peninsular India a general frequency of hourly readings of stage is usually
acceptable. For a few small upland catchments used for special purposes or research, 15 or 30
minute observations may be used to define the rainfall-runoff response. For the design of minor
irrigation schemes and bridge and culvert design on small catchments, 15 minute observations may
also prove useful.
Different practice will be adopted depending on whether measurement is by staff gauge only, chart
recorder or digital water level recorder, as summarised in Table 5.1.
For digital recorders the standard practice should be for a maximum time interval between readings of
one hour. On large natural lowland rivers, such an interval may be unnecessarily small, but few such
rivers in India are natural. Levels may change comparatively quickly as a result of river regulation and
abstraction a short distance upstream. Information on such changes is often helpful in scheme
operation and in analysis for example of times of travel. For small catchments, particularly those in
mountainous, high intensity rainfall areas a frequency of 15 minutes may be required.

Observation by Frequency Remarks
DWLR 15 min/hourly Dependent on size of catchment and
purpose for which data is required.
AWLR Hourly Depends on scale of chart, more frequent
readings (15 min.) could be extracted from
daily and/or strip charts
Staff gauge only Hourly
2 or 3 per day
Monsoon
Lean season
Staff gauge with AWLR or DWLR 2 or 3 per day
Stilling well inside reference level Daily
Table 5.1: Recommended observation frequency for stage measurements
For chart recorders, the record is of course continuous and information may be extracted at the
interval required. The ease of extraction will depend on the scale and size of the chart. However, it is
recommended that whenever possible a frequency of at least one hour is applied.
For manned stations with staff gauges only, hourly readings through a full 24 hr day (24/day) will
apply during the monsoon, with the season defined according to the local climate. During the lean
season, with the record already intermittent, hourly readings seem unjustified. Two or three readings
per day will be sufficient, with the provison, that in the event of unseasonal rainfall and river rise, the
observations are intensified and extended over the full 24 hours. In some circumstances one reading
a day might suffice.
For stations where the staff gauge is supplementary to the DWLR or AWLR, the staff gauge will be
read 3 times daily whilst the recorders are operating correctly but will otherwise revert to the practice
noted above.
Where auxiliary/secondary gauges exist they will be read at the same time intervals as the reference
gauge and the readings should be taken as close in time as possible.
For all stations level measurement should persist throughout the year, so long as there is flow and the
no-flow condition will be routinely observed and recorded daily. The latter observations are very
important; a nil flow is an observation, which must be recorded. Failure to do so results in confusion
between ‘no flow’ and missing data.
5.3 CURRENT METER MEASUREMENT FREQUENCY
The required frequency of current meter measurement at a stage-discharge site depends primarily on
the stability of the control section, as this will define how frequently gaugings are required to achieve a
given level of accuracy. The minimum number of gaugings required establishing a good stage-discharge
relationship for a stable, sensitive control is of the order of 10 -12 over the full flow range.
The nature of rivers in Peninsular India is such that the controls are often insensitive and the
uncertainties in current meter gauging are larger than desirable (> +/- 10%). Therefore considerably
more gaugings might be required in order to define the stage-discharge relationship. A minimum
number of 20 gaugings should at least be observed. The existing level of calibration is also important.
A precise interval between gaugings cannot be specified as the need to gauge may depend on the
occurrence of flow in a particular range.
Unstable channels and those affected by backwater or hysteresis resulting from unsteady flow will
require more persistent and frequent measurement than stable controls.
Recommended frequencies are proposed and these are summarised in Table 5.2. However, these
are only indicative and merely provided as a guide.

If more than one condition exists at a station then the condition requiring the most frequent gauging
should be applied.
If a change in the control is detected at any station or if current meter gauging suggests that the rating
has shifted, then gauging should be intensified until the new rating is defined throughout the range. At
sites with very unstable controls it might be necessary to derive a new stage-discharge relationship for
each season.
Mobile teams
Stage-discharge sites with reliable, stable ratings, which only require periodic current meter gaugings,
could be serviced by mobile field teams. Groups of sites should be selected within the same
area/locality, which could be visited by mobile teams to undertake the current meter gaugings.
Whenever feasible, the use of mobile teams should be encouraged, since it could reduce recurrent
use and make more efficient use of limited, skilled manpower. In addition, this would minimise the
amount of equipment required, since the teams could carry the current meters and accessories from
site to site.
Before each monsoon season the Executive Engineer will draw up a schedule of stations within his
jurisdiction outlining his recommendations where priority gauging is required, with ranges, where there
is currently insufficient gaugings for accurate definition of the stage-discharge relationship. The
schedule will be circulated to mobile and static teams for their action.
Station control Frequency Remarks
All stations (excluding structures) -
initial calibration
Daily, more frequent gauging if
appropriate to capture data for as wide a
range of events as possible
> 20 each in low, medium and high
flow range
Stable natural channel - well
calibrated
Monthly plus at least one high flow event a
year
The monthly gaugings can coincide
with routine chart changing/DWLR
downloading.
Backwater affected Daily if backwater source not known
Otherwise weekly
If an additional set of gauge posts
are installed or an additional
AWLR/DWLR, the changes in
surface water slope can be
estimated
Unstable channels with silt sand or
gravel
Daily or more frequent to obtain data for
high events
Intermittently unstable channels with
cobbles or boulders
Daily during monsoon
Weekly during lean season
Unsteady flow with looped rating Weekly Assumes that rate of change from
stage records of stage can be well
enough defined
Structures - Initially 6 gaugings over full modular
measurement range to confirm calibration
(performance) of structure, approx. 6
readings in the non-modular range.
Including 2 low flows. The modular
limit should be defined
Structures - after initial calibration
(performance) check
1 - 2 gaugings per year within modular
range, 3 - 4 gaugings a year in non-modular
range
Table 5.2: Recommended observation frequency for current meter gauging

Hydrometry: Design Manual Volume-4 2003

Hydrometry: Design Manual Volume-4 2003

Recommandé

Recommandé

Contenu connexe

Tendances

Tendances (20)

En vedette

En vedette (17)

Similaire à Hydrometry: Design Manual Volume-4 2003

Similaire à Hydrometry: Design Manual Volume-4 2003 (20)

Plus de indiawrm

Plus de indiawrm (20)

Dernier

Dernier (20)

Hydrometry: Design Manual Volume-4 2003